dave chan_ international conference on soft soil engineering. _4, 2006, vancouver, british...

SOFT SOIL ENGINEERING

2007 by Taylor & Francis Group, LLC

BALKEMA Proceedings and Monographsin Engineering, Water and Earth Sciences


http://www.crcnetbase.com/action/showImage?doi=10.1201/9781439833926.fmatt&iName=master.img-000.jpg&w=77&h=78

PROCEEDINGS OF THE FOURTH INTERNATIONAL CONFERENCE ON SOFT SOILENGINEERING, VANCOUVER, CANADA, 46 OCTOBER 2006

Soft Soil Engineering

EditorsDave ChanDepartment of Civil & Environmental Engineering, University of Alberta,Edmonton, Alberta, Canada

K. Tim LawDepartment of Civil and Environmental Engineering, Carleton University,Ottawa, Canada

LONDON / LEIDEN / NEW YORK / PHILADELPHIA / SINGAPORE


Copyright 2007 Taylor & Francis Group plc, London, UK

All rights reserved. No part of this publication or the information contained herein may be reproduced,stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, byphotocopying, recording or otherwise, without written prior permission from the publisher.

Although all care is taken to ensure the integrity and quality of this publication and the information herein,no responsibility is assumed by the publishers nor the authors or editors for any damage to property orpersons as a result of operation or use of this publication and/or the information contained herein.

Published by: Taylor & Francis/BalkemaP.O. Box 447, 2300 AK Leiden, The Netherlandse-mail: [email protected]/engineering, www.crcpress.com

ISBN13: 978-0-415-42280-2

Printed in Great Britain


www.taylorandfrancis.co.ukwww.crcpress.comwww.crcpress.comwww.taylorandfrancis.co.uk

Soft Soil Engineering Chan & Law (eds) 2007 Taylor & Francis Group, London, ISBN13 978-0-415-42280-2

Table of Contents

Preface XI

Organizing Committee XIII

Keynote papers

Stability analysis accounting for macroscopic and microscopic structures in clays 3K.Y. Lo & S.D. Hinchberger

Soft soil stabilisation with special reference to road and railway embankments 35B. Indraratna, C. Rujikiatkamjorn, V. Wijeyakulasuriya, M.A. Shahin & D. Christie

Modelling and numerical simulation of creep in soft soils 57P.A. Vermeer, M. Leoni, M. Karstunen & H.P. Neher

Experimental study on shear behavior and an improved constitutive model of saturated sandunder complex stress condition 73M. Luan, C. Xu, Y. He, Y. Guo, Z. Zhang, D. Jin & Q. Fan

Embankment and dams

Sensitivity analysis of magnetic extensometers for measuring vertical movementof earth dams on soft soils 95R.J. Chenari

Building an embankment with simultaneous vacuum loading 105B.T. Wang & K.T. Law

Failure of a column-supported embankment over soft ground 117W.M. Camp III & T.C. Siegel

Performance of highway embankments on Bangkok clay 123S. Apimeteetamrong, J. Sunitsakul & A. Sawatparnich

Geogrid-reinforced roadway embankment on soft soils: A case study 129R. Vega-Meyer, R.S. Garrido, A.R. Piedrabuena, I.N. Larios & R.P. Lapuente

Monitoring the staged construction of a submerged embankment on soft soil 139W.F. Van Impe, R.D. Verstegui Flores, J. Van Mieghem, A. Baertsoen & P. Meng

Optimal design of grillage supporting structures for stabilizing slopes 145Y. Zhu & Y. Zhou

Measured settlements of the Srmin high embankment 153P. vanut, M.R. Turk & J. Logar

Joint calculation of a foundation and soil of the large-scalestructure in view of creep 159S. Aitalyev, N. Ter-Emmanuilyan, T. Ter-Emmanuilyan & T. Shmanov

Foundation

Pile resistance variations over time for displacement piles in young alluvium 171A.A. Hanifah, M.N. Omar, N.F.A. Rahman & T.K. Ong

V


Group effect on model piles under axial monotonic loading 179A.L. Kouby, J. Canou & J.C. Dupla

2D numerical modeling of Pile-net composite foundation of high-speedrailway embankment in soft soils 189J.-D. Niu, L.-R. Xu, B.-C. Liu & D.-W. L

Study on the influence of pile foundation due to excavation 199Y. Zhang, J. Zai & K. Qi

Study on long-term settlement behavior of driven pile foundation in soft soil 205H.-B. Zhou, Z.-C. Chen & N.-F. Hong

Analyzing the static tests of boring piles through CFA technology 213A.Zh. Zhusupbekov, Y. Ashkey, V.N. Popov, A.J. Belovitch & G.A. Saltanou

Large scale experiment and case study

Design and performance of a combined road-channel-dike structure founded on verysoft Bangkok clay 219P. Boonsinsuk

Improvement of a very soft dregded silty clay at the port of Valencia (Spain) 231M. Burgos & F. Samper

Study of geosynthetic reinforced subgrade expressway in Taiwan 237S.-J. Chao

A study on dynamic shear modulus ratio and damping ratio of recently deposited soilsfor southern region of Jiangsu province along Yangtze River, China 245G.-X. Chen, X.-Z. Liu & D.-H. Zhu

Investigations on improvement of soft ground treated by various vertical drainsunder embankment on soft clay foundation 251H.I. Chung & J. Yu

Static and seismic stability of geogrid reinforced-soil segmental bridge abutmentsconstructed on soft-soil 257K. Fakharian & I.H. Attar

Geotechnical behavior of organic soils of North Sarawak 267S.R. Kaniraj & R.R. Joseph

Behavior characteristics of unreinforced and reinforced lightweight soils 275Y.-T. Kim & H.-J. Kim

A case study of building damage risk assessment due to the multi-propped deep excavationin deep soft soil 281S.-J. Lee, T.-W. Song, Y.-S. Lee, Y.-H. Song & J.-K. Kim

Study on jackup spudcan punch-through 291C.F. Leung, K.L. Teh & Y.K. Chow

Apparent earth pressure of soft soils overlying hard bedrock at South Link in Stockholm 299J. Ma, B.S. Berggren, P.-E. Bengtsson, H. Stille & S. Hintze

Performance of stone column encased with geogrids 309S.N. Malarvizhi & K. Ilamparuthi

Strength distribution of soft clay surround lime-column 315A.S. Muntohar & J.-L. Hung

The investigation of mud tailings and a comparison of different test methodswith 3rd world constraints 321W. Orsmond

VI


A geotechnical data base development and applying data mining techniques to extractthe common trendes of offshore geotechnical properties of South ParsGas Field/Persian Gulf IR-IRAN 327H. Shiri GJ. & M.H. Pashnehtala

Inaccurate interpretation of offshore geotechnical site investigation results and riskassociated: A case study of conductors collapse in driving 333H. Shiri GJ. & B. Molaei

Behaviour of plate anchors under short-term cyclic loading 341S.P. Singh & S.V. Ramaswamy

Dissipation process of excess pore water pressure caused by static pressed pile in soft soil 347W. Wang, J. Zai & T. Lu

Research on control of settlement and stabilization of high subgrade besidehill and above soft foundation in Wenzhou expressway 351H.-L. Yao, Y.-Q. Zhou, Z. Lu & Q. Zhou

Engineering performances of soil disturbed by underground mining and its application 357G.-Y. Yu, P. Sheng & L.-B. Wang

Material behaviour

Compressibility properties of reconstituted organic soils at Khulna Region of Bangladesh 367M.R. Islam, M. Alamgir & M.A. Bashar

Temperature effects on engineering behaviour of soft Bangkok clay 373D.T. Bergado, H.M. Abuel-Naga & A. Bouazza

Discharge capacity of vertical drains installed in soft ground with timeby laboratory small and large tests 381H.I. Chung, Y.S. Lee & Y.M. Park

Study of preconsolidation pressure values derived by the modified Casagrande method 385I.N. Grammatikopoulos

Modeling sand behavior in constant deviatoric stress loading 389R. Imam & N. Morgenstern

Identification of a general poro-viscoelastic model of one-dimensional consolidating soft soil 397C.J. Leo

Characterisation of peat using full flow penetrometers 403N. Boylan & M. Long

Experimental study of ageing effect on the undrained shear strength of silty soil 415M. Ltifi

New relationships to find the hydraulic conductivity and shear wave velocity of soft Pusan clays 421K.G. Rao & M. Suneel

Geotechnical characteristics of a very soft dredged silty clay and a soil-cement mixin Valencia Port (Spain) 427M. Burgos & F. Samper

Consolidation behavior of a soft clay composite 437A.P.S. Selvadurai & H. Ghiabi

Laboratory testing of a soft silty clay 447H. Ghiabi & A.P.S. Selvadurai

Effect of heating on pore water pressure of soft bentonite 457A.N. Sinha & O. Kusakabe

VII


Undrained strength and compressibility of mixtures of sand and coal 463C. Stamatopoulos & A. Stamatopoulos

The assessment of destructuration of Bothkennar clay using bender elements 471J. Sukolrat, D. Nash, M. Lings & N. Benahmed

Softening characteristics of soil cement on the condition of soaking 481C.-J. Yin, X.-H. Wang & S.-C. Ma

Progressively destructurated undrained strength of natural soils 485C. Zhou

Numerical modelling and theoretical development

Numerical modeling of interaction between flexible retaining wall and saturatedclayey soil in undrained and drained conditions 493A.M. Bazrafshan & A. Pak

Numerical modeling for ground settlement due to two-tunnel shielding construction 499Y. Bian, F. Zhuo, Y. Zhu & X. Ji

Numerical modeling of an embankment on soft ground improved by vertical rigid piles 505O. Jenck, D. Dias & R. Kastner

Numerical simulation of passively loaded piles adjacent to embankment constructedon soft Bangkok clay 515R. Katzenbach & S. Pokpong

Three dimensional nonlinear finite element analyses for horizontal bearing capacityof deeply-embedded large-diameter cylindrical structure on soft ground 521Q. Fan, M. Luan & Q. Yang

Numerical modelling of a very soft dredged silty clay improvement in Valencia port (Spain) 531F. Samper & M. Burgos

Slope stability and landslide

A new method for slope stability analysis of foundation pit due to groundwater seepage 541G. Chen, C. Li & Y. Fan

A simplified method for stability analysis of reinforced embankments 547Y.H. Chen, T. Zhang, X.H. Ma, Y.Q. Zhou, M.J. Gao & C.C. Gu

Stability analysis of expansive soil slope and its slope remedeations 553R.Q. Huang & L.Z. Wu

Determination of non-circular critical slip surface by harmony search algorithm inslope stability analysis 557L. Liang, C. Shichun & Y.M. Cheng

Simple critical state model predicting the response along slip surfaces 563C. Stamatopoulos

The other soil parameters in stability limit analysis of soil-nailed walls in soft soil 573Y. Yang

Soil improvement

Centrifuge study on assessment of geological barrier of soft ground with floating type sand drains 581B.L. Amatya, J. Takemura, T. Ashida & O. Kusakabe

The use of dynamic compaction in liquefaction hazards mitigation at reclaimed lands in Assalouyehpetro-chemical complex-Iran 587S.S. Yasrobi & M. Biglari

VIII


Optimization of strength and ductility of Class C fly ash stabilized soft subgrade soils 595S. Bin-Shafique, A. Senol, C. Benson & T. Edil

Stabilization of soft clay site for development using Rammed Aggregate PiersTM 601W. Sheu, E.M. Vlaeminck, B.T. FitzPatrick & J. Bullard

Improvement of soft soils by static SCP using a hydraulically-operated rotary penetration 611R. Shiozaki, K. Uehara, S. Ikenoue, K. Ookori, Y. Umeki, M. Mori & M. Fukue

Promotion of consolidation for dredged soft sediments using permeable bags 619M. Fukue, K. Kita, C. Mulligan, K. Uehara, Y. Umeki & T. Inoue

Estimation of the settlement of improved ground with floating-type cement-treated columns 625R. Ishikura, H. Ochiai, K. Omine, N. Yasufuku & T. Kobayashi

Improvement for soft soil by soil-cement mixing 637S. Jaritngam & S. Swasdi

Improving engineering properties of soft clayey soils using electrokinetics: A laboratorybased investigation 643S. Jayasekera & S. Hall

3D modelling of deep mixing 649H. Krenn, M. Karstunen & A. Aalto

Trafficability evaluation of PTM treated dredged soil deposit 657S.-R. Lee, W.-Y. Byeon, H.-G. Park & S.-H. Jee

Comparison of performance between the dry and wet Deep Mixing method in softground improvement 667S. Liu, L. Chen & Y. Deng

A fundamental study on the remediation of contaminated soil with heavy metals basedon electrokinetic and magnetic properties 673K. Omine, H. Ochiai & N. Yasufuku

Effect of zeolite and bentonite on the mechanical properties of cement-stabilized soft clay 681A.A.-M. Osman & A. Al-Tabbaa

Enhancement of strength of soft soils with fly ash and lime 691P.V. Sivapullaiah, B. Katageri & R.N. Herkal

Soil improvement using compaction grouting a laboratory investigation on the confiningpressure and injection rate in completely decomposed granite 697S.Y. Wang, D. Chan, K.C. Lam, S.K. Au & L.G. Tham

The physical and mechanical properties of lime stabilized high water content expansive soil 703B. Wang, X. Ma, W. Zhang, H. Zhang & G. Chen

Combined preloading compaction and composite ground to treat the soft subgrade of highway 709G. Zheng, S. Liu & H. Lei

Theoretical analysis and constitutive modelling

The use of statistic analysis in predicting of ground and wall movements in soft clay 717P. Chaichi & N. Shariatmadari

Analytic solutions of consolidation of fine-grained compressible soils by vertical drains 723C.J. Leo

Improved stress-strain model of soft soil based on energy dissipation theory 731T. Lu & W. Wang

Application of BP neural network in identifying soil strata by CPTU 735S.-Z. Ma, H.-B. Jia, G.-T. Meng & S.-L. Liu

IX


A simplified plastic hysteretic model for multi-directional nonlinear site response in soft soils 741J.M. Mayoral, J.M. Pestana, M.P. Romo & R.B. Seed

Mathematical description of consolidation test 749G.-X. Mei, J.-M. Zai & J.-H. Yin

Back analysis of three case histories of braced excavations in Boston Blue Clayusing MSD method 755A. Osman & M. Bolton

Effect of ratio of influence zone and type of vertical drain on consolidation of soft claydue to radial flow 765A.V. Shroff, M.V. Shah, T. Khan & N. Joshi

Study on the depth of crack propagation of unsaturated expansive soils 775Q. Yang, P.-Y. Li & M.-T. Luan

Elastic viscoplastic modeling of two cases involving PVD improved Hong Kong marine clay 779Z. Fang, J.H. Yin, C. Zhou & J.G. Zhu

X



Preface

Soft soil is found in many places in the world and especially in coastal cities like Shanghai,Tianjin andVancouver.Soft sensitive clay, such as the Quick Clay along the St. Lawrence Seaway and in the Ottawa region in Canada,provides many challenges to geotechnical engineers when building in or on this material. In many instances, softsoil has to be treated using a variety of soil improvement techniques to improve its strength, deformation andhydraulic properties.

The Fourth International Conference on Soft Soil Engineering provided an opportunity for geo-professional,geotechnical engineers, academic and researchers, to share their experiences and research results on soft soils.It was a continuation of previous three conferences held in Guangzhou, Nanjing and Hong Kong. The FourthInternational Conference on Soft Soil Engineering was held in Vancouver where there are soft soil problemssince Vancouver is situated at the river delta of the Fraser River. Delegates from over 20 countries gathered inHotel Vancouver between October 4 and 6, 2006 to discuss soft soils engineering. The conference dealt withmany technical issues of soft soil engineering such as soft soil construction, ground improvements, constitu-tive behaviour of soft soils, numerical modeling, hazard mitigation and post hazard ground investigation andimprovements. There were four keynote lectures given by leading professors/engineers from Canada, Germany,Australia and China who shared their research findings and experiences in dealing with soft soils.

XI



Organizing Committee

Conference ChairProf. Dave Chan

Steering CommitteeProf. D. H. Chan Prof. C. F. LeeDr. C. K. Lau, Prof. L. G. ThamProf. K. T. Law Prof. J.H. Yin

International Advisory CommitteeProf. Dave Chan Dr. H. L. LiuDr. Dennis Becker Dr. Charles NgDr. Dennis Bergado Prof. Pieter VermeerProf. Buddhima Indraratna Prof. Richard WanDr. Suzanne Lacasse Dr. H. S. YuDr. K. C. Lam Prof. Askar ZhusupbekovProf. Serge Leroueil

Local Organizing CommitteeDr. Ranee Lai (Chair) Mr. Makram SabbaghDr. Reza Iman Mr. Daniel YangMr. Gavin Lee Dr. Mustapha ZerguonMr. Howard Plewes

Technical Program CommitteeProf. Tim Law (Chair) Prof. Julie ShangProf. Masaharu Fukue Prof. Siva SivathayalanDr. Kai Sing Ho Prof. Keizo UgaiProf. Jean-Marie Konard Prof. Baotian WangProf. Maotian Luan Dr. Quentin Yue

Organized byThe University of AlbertaThe University of Hong KongThe Hong Kong Polytechnic University

Supported byThe Canadian Geotechnical Society

XIII


Keynote papers



Stability analysis accounting for macroscopic and microscopicstructures in clays

K.Y. Lo & S.D. HinchbergerGeotechnical Research Centre, Faculty of Engineering, The University of Western Ontario, London,Ontario, Canada

ABSTRACT: Geotechnical Engineering has advanced to the present stage that various types of earth structurescan be designed and constructed safely and economically in most instances. However, in some cases, difficultyarises either in the form of failure during construction or after many years in existence. The soils in which theseproblems occur include but are not limited to highly sensitive clays and stiff fissured clays of various geologicalorigins. These clays possess pronounced macroscopic and microscopic structures that control the strength anddeformation properties. Macroscopic structures are visible features that include fissures, joints, stratificationsand other discontinuities in an otherwise intact soil mass. Microscopic structures would include soil fabricand cementation bonds. A typical soft clay deposit usually is composed of a weathered crust at the top that isfissured and thus macroscopic structures are dominant and soft clay at depth in which microscopic structuresare significant. The properties of these clays are complex, having a stress-strain relationship that exhibits a peakstrength and a post peak decrease in strength, a non-linear failure envelope, strength anisotropy and a significantdecrease in strength with a slower rate of testing or longer time to failure.

This paper explores the implications of microscopic and macroscopic structure on stability problems and theconditions under which difficulties arise. Results of laboratory and field tests together with case histories showthat the dominant effect of a macroscopic structure is exhibited in the reduction of undrained and drained strengthwith the sample size. The mass strength, whether in the undrained or drained condition, is only a fraction of theintact strength. Design analysis for stability conditions should therefore start with the mass strength at initialtime followed by a reduction in strength as time progresses. A case history of an embankment founded on stifffissured clay on which it failed after 32 years is analyzed in detail to illustrate progressive development of plasticzones with construction details and time. The effect of cementation bonds in influencing the strength propertiesof soft clays is studied by artificially deposited bonds using the electro-kinetic process and examination with theelectronic microscope. It is shown that in addition to the classical increase in strength with decrease in watercontent, a strength increase occurred with time due to the deposition of cementation bonds by diffusion. Animportant bonding agent is identified and its effect on bond strength is compared with bonding in natural clays.As the height of an embankment founded on a sensitive clay deposit is increased, a plastic zone will develop andincrease in size. The pore pressures at a point will increase at a greater rate when the point is engulfed by theplastic zone as a result of bond breakage. Concurrently, the strength will drop to the post-peak state. Case historiesof embankments on these clays are analyzed to illustrate the propagation of the plastic zone in controlling thefoundation behaviour at imminent instability. The difference in performance of embankments with differentgeometries in the same clay deposit is investigated. It is shown that the stability and subsequent strength changesare controlled by the loading geometry and extent of the plastic zone. Finally, design considerations are suggestedto accommodate the effects of the macroscopic and microscopic structures in these clays.

1 INTRODUCTION

At present, soft clay engineering has advanced to thestage that earth structures can be designed economi-cally and constructed safely in most cases. There are,however, circumstances in which failure has occurredduring construction or after many years in existence inspite of the detailed field and laboratory investigations

that had been carried out. The soils in which theseproblems occurred include but are not limited to stifffissured clays and highly sensitive clays, as exem-plified by the following two well-documented caserecords.

The first case involved an embankment constructedat Nanticoke, Ontario, on a deposit of stiff fissuredclay after extensive field and laboratory investigations.

3


Table 1. Properties of some clays.

UndrainedSite LL PI LI Strength (kPa) Sensitivity References

Nanticoke 55 31 0.06 380 1 Lo et al. (1969)Wallaceburg (depth 4.2 m) 46 18 1.0 37 6 Becker (1981)Sarnia Till 38 26 0.16 150 2 Quigley and Ogunbadejo(1976)St. Vallier 60 37 0.97 43 20 La Rochelle and Lefebvre (1970)St. Louis 50 23 1.83 43 50 La Rochelle and Lefebvre (1970)St. Alban 50 23 2.4 11 14 La Rochelle et al. (1974)Olga 60 32 1.6 10 13 Dascal et al. (1972)Vernon 65 40 1.14 30 4 Crawford et al. (1995)

The embankment was originally designed for a max-imum height of 17 m (locally) with 2:1 slopes. It wasconstructed in 1969 as a containment dyke for flyash disposal. Surficial instability occurred at variousperiods after construction with time to failure of sev-eral months to several years. The downstream slopewas flattened in 1977 to 2.75:1. However, instabilityoccurred at 32 years after construction.

The second case involved a dramatic and mostinstructive case record presented by Crawford et al.(1995) who described two consecutive failures of anembankment on soft clay, in spite of the fact that twotest embankments were already constructed on eitherside of the failures and that the test embankments werehigher than the embankments that failed.

The conditions under which these problemsoccurred are explored in this paper. Additional con-siderations to conventional design methodology aresuggested.

2 BEHAVIOUR OF INTACT CLAYS

Highly sensitive clays and stiff fissured clays of variousgeological origins possess pronounced macroscopicand microscopic structures that control the strengthand deformation properties.

Macroscopic structures are visible features thatinclude fissures, joints, stratifications and other discon-tinuities in an otherwise intact soil mass. Microscopicstructures would include soil fabric and cementationbonds identifiable, for example, using electron micro-scope techniques. A typical soft clay deposit usually iscomposed of a weathered crust at the top that is fissuredand thus macroscopic structures are dominant and softclay below the crust wherein microscopic structuresare significant.

The properties of these clays are complex, having astress-strain relationship that exhibits a peak strengthand a post peak decrease in strength, a non-linearfailure envelope, strength anisotropy and a significantdecrease in strength with a slower rate of testing orlonger time to failure. Leroueil (2005) has presenteda comprehensive review of the behaviour of sensitive

Figure 1. Effective strength envelope of Nanticoke Clayfrom 4.5 m depth.

clays. The following discussion covers the behaviourof relatively insensitive intact stiff clays, and a fewadditional observations are also made on the behaviourof sensitive clays. In order to avoid the effects of sam-ple disturbance, only results of tests from specimenstrimmed from block samples or high quality largediameter samples are considered.

2.1 Non-Linearity of Mohr-Coulomb envelope

Traditionally, engineers have adopted a linear relation-ship for the Mohr-Coulomb failure envelope. In reality,test results have invariably shown that the envelope isintrinsically nonlinear. However, the details of nonlin-earity are markedly different between highly sensitiveand relatively insensitive clays. Properties of the claysdiscussed in the following paragraphs are shown inTable 1.

Figure 1 shows the Mohr-Coulomb envelope deter-mined from intact specimens trimmed from blocksamples of insensitive stiff fissured clay taken at theNanticoke Generation Station, Ontario (Valle 1969).It can be seen that the envelope is mildly nonlinear overa wide stress range with the strength increasing witheffective stresses. This behaviour is also exhibited inother materials such as intact rock and concrete.

Figure 2 shows the results of tests on WallaceburgClay (Becker 1981) near Sarnia, Ontario. The clay is

4


http://www.crcnetbase.com/action/showImage?doi=10.1201/9781439833926.pt1&iName=master.img-000.jpg&w=191&h=121

Figure 2. Effective strength envelope of firm WallaceburgClay. (after Becker 1981)

Figure 3. Stress condition at failure and stress paths, CIUtests at i= 0 on St. Louis Clay. (after Lo and Morin 1972)

firm at the depth of testing with a liquidity index ofabout one and a sensitivity of six by field vane tests.It can be seen that the envelope is nonlinear. However,over the stress range from 70 kPa to 90 kPa straddlingthe preconsolidation pressure, there is little increase instrength with increasing effective stress.

Figure 3 shows the results of CIU tests on St. LouisClay (St = 50) (Lo and Morin 1972). The envelopeis strongly nonlinear. The remarkable feature is thatthere is a significant decrease in strength with anincrease in effective stress as the consolidation pres-sure approaches the preconsolidation pressure. Similarbehaviour can be seen for St. Vallier Clay (St = 20) inFigure 4.

2.2 Anisotropy

The results of triaxial compression tests on specimensfrom St. Vallier with their axes trimmed at i= 0, 45and 90 from the vertical are shown in Figure 4. Theapparent anisotropy of the strength envelope is evi-dent although the trend of decreasing strength withan increase in effective stress is less distinct. Thedecrease in strength with an increase in effective stressmay be attributed to bond breakage, a progressiveprocess of damage to the microscopic soil structure

Figure 4. Results of CIU tests at different i on St. VallierClay. (after Lo and Morin 1972)

Figure 5. Stress-strain relationship of St. Vallier Clay fromdrained triaxial tests. (after Lo 1972)

in which the strength loss due to bond breakageovershadows the strength gain due to effective stressincreases until most of the bonds are broken, where-upon their effects are obliterated. At effective stressesthat exceed the preconsolidation pressure, the enve-lope enters into the unstructured portion where thestrength increases linearly with effective stress. Fur-ther study of cementation bonds will be discussed inSection 5.

2.3 Effect of time

The stress-strain relationship of specimens from blocksamples of St. Vallier Clay measured in isotropicallyconsolidated drained triaxial tests at consolidationpressures below the vertical preconsolidation pressureare shown in Figure 5 (Lo 1972). One series of testswas performed at the conventional axial stain rate of

5


http://www.crcnetbase.com/action/showImage?doi=10.1201/9781439833926.pt1&iName=master.img-001.jpg&w=191&h=117http://www.crcnetbase.com/action/showImage?doi=10.1201/9781439833926.pt1&iName=master.img-002.jpg&w=191&h=104http://www.crcnetbase.com/action/showImage?doi=10.1201/9781439833926.pt1&iName=master.img-003.jpg&w=191&h=97http://www.crcnetbase.com/action/showImage?doi=10.1201/9781439833926.pt1&iName=master.img-004.jpg&w=157&h=215

Figure 6. Effect of strain rate on the peak strength ofSt. Vallier Clay (after Lo and Morin 1972)

0.1% per hour while the other series was performed40 times slower. It may be seen that both strength andstiffness decreased with the slower rate of testing. Thedependence of the failure envelope on the rate of test-ing, including CIU tests, are shown in Figure 6 (Lo andMorin 1972). Since the strain rates of laboratory testsare vastly different from the strain rates in the field,the results of these tests indicate that the effect of timeto failure is a significant factor to be considered in thedesign of earth structures.

2.4 Post-Peak envelope

It has often been found that in most natural soils, thestrength decreases after the peak strength has beenreached. For sensitive clays, it was recognized that anenvelope defined by the state of stresses at strains in theorder of 6% to 10% is of particular engineering signif-icance for the analysis of slope stability (Lefebvre andLa Rochelle 1974, Lo and Morin 1972). It was con-sidered that the effect of anisotropy, time rate and thepotential for progressive failure all tend to reduce thepeak strength envelope towards the post-peak strengthas shown in Figure 7 (for details see Lo and Morin1972). Analyses of natural slope failures in ChamplainClays showed that the results lie close to the post-peakenvelope as shown in Figure 8 (Lo and Lee 1974). Forfirst time slides of cut slopes, the results lie above thepost-peak envelope (see points for Orleans (Lo 1972),Lachute 1 and Lachute 2 (Lefebvre 1981)) as expected,since the progression of progressive failure can satisfythe limiting equilibrium condition before the post-peakstrength is reached over the entire slip surface.

An important contribution to the verification ofthe concept of the post-peak strength was made byLaw (1981). A comprehensive series of tests on speci-mens prepared from 100 mm Osterberg samples fromRockcliffe in the Ottawa region was performed usingdifferent stress paths. The results showed that:

(a) The brittleness of both a sensitive clay and a stiffclay decreases from a constant 3 test to constant

Figure 7. Influence of different physical factors on stressconditions at failure in sensitive clays. (after Lo and Morin1972)

Figure 8. Summary plot for natural slope failures in Cham-plain Sea Clay. (after Lo and Lee 1974, with additionalcases)

p test to constant 1 test. However, the brittlenessof both clays is still manifested (see Figure 9).

(b) The post-peak envelope is independent of the stresspath and is remarkably similar to that deduced byLo and Lee (1974) (see Figure 10).

It appears, therefore, that the concept of post-peakenvelope remains valid since its inception as a basisfor the evaluation of the stability problem.

3 THE MASS STRENGTH OF FISSUREDCLAYS

In soil deposits that are essentially free of discontinu-ities, the properties of intact specimens measured in thelaboratory would be representative of field behaviour,apart from accounting for their complex behaviour.In a soil mass populated by features such as fissuresand joints, the properties measured in small intactspecimens in conventional sampling and testing canbe misleading.

6


http://www.crcnetbase.com/action/showImage?doi=10.1201/9781439833926.pt1&iName=master.img-005.jpg&w=191&h=122http://www.crcnetbase.com/action/showImage?doi=10.1201/9781439833926.pt1&iName=master.img-006.jpg&w=191&h=97http://www.crcnetbase.com/action/showImage?doi=10.1201/9781439833926.pt1&iName=master.img-007.jpg&w=191&h=164

Figure 9. Effect of stress path on brittleness index of clays.(after Law 1981)

Figure 10. Summary of results of stress path tests onRockcliffe Clay. (after Law 1981)

It has been recognized that the macroscopic struc-tures of a clay can dominate its strength behaviour andthat the strength of the soil mass is only a fractionof that of the intact material (e.g. Bishop and Lit-tle 1967, Lo 1970). Macroscopic structures includefissures, joints and other discontinuities in an other-wise intact soil mass. For comparison, the effectivestress parameters of some stiff clays in the intact state,along natural surfaces of weakness, and in the resid-ual state are given in Table 2. It may be seen thatthe strength along the discontinuities is much lowerthan the intact material but distinctly higher than theresidual strength.

Many hypotheses for the mechanisms of formationof discontinuities in clays have been put forward thatinclude but are not limited to:

(a) weathering: one of the generally accepted mecha-nisms, including cycles of deposition, desiccation,erosion and redeposition;

(b) syneresis: the colloidal process in which particlesare drawn together, forming honeycomb patternsof cracking during aging;

(c) one-dimensional swelling due to removal of over-burden such that the strain required to reachpassive failure is attained (Skempton 1961);

(d) tectonic stresses;(e) stress relief and valley rebound due to erosion;(f) slumps on steep rock valleys during deposition,

forming large scale discontinuities;(g) glacial shear;(h) temperature effects.

While joints, shear zones and faults affect thedirectional stability of an earth structure, the mostubiquitous discontinuities are fissures prevalent in stifffissured clays and the crust of soft or firm clay deposits.

An example of the large difference in undrainedstrength between fissures and intact material of Nanti-coke Clay is shown in Figure 11. Because of the largedifference in strength, whether in the undrained (Fig-ure 11) or drained (Table 2) condition, the presence offissures considerably weakens the otherwise intact soilmass. The degree of weakening would depend on thedifference between the intact strength and the fissurestrength as well as the density and size distribution ofthe fissures. An example of a decrease in strength withsample size (area of potential failure surface) is shownin Figure 12.

The impact of macroscopic structures on the stabil-ity of earth structures such as cut slopes is substantial.Table 3 summarizes some case histories of failurein stiff fissured clays. The quantities Su and Su,mrepresent the strengths from conventional unconsoli-dated undrained tests and the mass strength from backanalysis of failure, respectively. Fu is the factor ofsafety computed from conventional U-U strength. Itcan be seen that these conventional factors of safetyconsiderably exceed one. It follows, therefore that adesign approach without consideration of macroscopicstructure could be unsafe.

4 UNDRAINED STRENGTH OF THE CRUSTSOF SOFT CLAY DEPOSITS

It is often found in soft or firm clay deposits thata stiffer crust exists of one to several metres thick.The crust is typically fissured with high vane strength.The strength decreases through the transition zoneand from there to the soft layer where the strengthincreases again (see, for example, Figure 28 and 40).The assumption of the value of undrained strength forthe crust has a significant effect on the design factorof safety for embankments on soft clays.

The field vane test is commonly used for the mea-surement of undrained strength in field investigations.However, the failure surface is cylindrical in the field

7


http://www.crcnetbase.com/action/showImage?doi=10.1201/9781439833926.pt1&iName=master.img-008.jpg&w=191&h=157http://www.crcnetbase.com/action/showImage?doi=10.1201/9781439833926.pt1&iName=master.img-009.jpg&w=191&h=146

Table 2. Strength of clays along discontinuities.

IndexProperties Strength Parameters

WL WP WN Type of Intact Discon-Clay % % % Discontinuities Material tinuities Residual Reference

c cw w cr rkPa () kPa () kPa ()

Nanticoke Clay 58 24 26 Fissure 31 36 13 18 7 15 Lo & ValleOntario (depth= 6 m) (1970)Upper Siwalik 60 28 16 Minor Shear 58 22 17 16 0 14 Skempton and PetleyClay Sukian (1967)Blue London 70 27 28 Joint and Fissure 31 20 7 18.5 1 16 Skempton et al.Clay Wraysbury (1969)Barton Clay 83 32 30 Fissure 26 38 6 18 3 13 Marsland & ButlerHampshire (1967) and Corbett (1967)Magho District Bedding Joint 8 25 6 18 5 8 Prior and Fordham (1974)Northern IrelandShale

Figure 11. Stress-strain relation of intact and fissure sam-ples-unconsolidated-undrained tests. (after Lo 1970)

vane test while fissures are approximately planar.Therefore, the likelihood of containing fissures in thevane test is small and the vane test measures essentiallythe intact strength (Lo 1970) apart from disturbance

Figure 12. Strength-size relation, Nanticoke Clay from 6 mdepth. (after Lo 1970)

during insertion and the effects of strength anisotropy.The effect of macroscopic structure therefore wouldrequire the field vane strength to be reduced to corre-spond to the mass strength of the crust. Field results ofcrust mass strength are scarce but the work of Quigleyand Ogunbadego (1976) and Lefebvre et al. (1987) arediscussed below.

4.1 Sarnia till

In a comprehensive study of the properties of Sar-nia Till in connection with pollutant migration in aSarnia landfill site, Quigley and Ogunbadejo (1976)performed large in situ shear box tests on the SarniaTillusing the same equipment and similar procedure as Loet al. (1969). The tests were performed at three depthsof 1.5, 3.0 and 4.5 metres. The first two levels corre-spond to the crust and the third level corresponds tothe transition zone below the crust. The results, shownin Table 4, indicated that the ratio of mass strengthto intact strength increases with depth, reflecting thedecreasing intensity of fissuring with depth. It is also

8



Table 3. Some case records of failure on fissured clays and rock.

Soil WL Wp WN Su Su,mCase Record Type Structure (%) (%) (%) IB kPa Fu kPa Reference

Bradwell 1 Brown Cut 95 30 33 0.05 97 1.8 54 Skempton and(England) London Clay La Rochelle (1965)Bradwell 2 Brown Cut 95 30 33 0.05 97 1.9 50(England) London ClayWravsbury Blue Cut 73 28 28 0.0 118 3.3 36 Simons (1967)(England) London ClayDurgapur Blue Cut 58 20 23.4 0.09 113 8.7 13 Dastidar (1967)(India) Silty ClayDunvegan Clay Fill 50 24 22 0.04 217 2.6 83 Hardy et al. (1962)(Alberta) ShaleSouth Clay Cut 80150 1827 1935 70 2.5 28 Peterson et al. (1960)Saskatchewan ShaleWitbank Coal Pillar 31 900 7.4 4300 Bieniawski (1968)Colliery (SouthAfrica)Houston (Texas) Fissured Anchored 65 22 22 0 97217* 2.2 2497 Daniel & Olson

Clay Sheet (1982)Pile Wall

Note: IB =Brittleness Index; Su =Undrained Strength from Conventional UU tests; Fu =Factor of Safety used on SuSu,m =Mass Strength Computed from Failure* Increases with depth

Table 4. Effect of fissures on the intact undrained strength of clays.

Soil Deposit Depth (m) Sui (kPa) Su,m (kPa)Su,mSui

Sarnia Till 1.5 (Crust) 280 55 0.203.0 (Crust) 250 104 0.414.5 (Transition) 150 85 0.56

Olga Sensitive Clay 0.21.2 (Crust) 75 (25) 18 0.24 (east trench)80 (40) 18 0.23 (north trench)

Nanticoke G.S. 3.3 333 56 0.17Fissured Clay 4.8 390 95 0.24

6.1 371 97 0.26Brown London Clay, Maldon 1.42.0 77 31 0.40

Note: Sui = intact undrained strength from UU tests or vane testSu,m = undrained mass strength from in situ shear box tests

interesting to note that there is very little post-peakdrop in strength for this clay from the in situ shear boxtest with the brittleness index being about 0.07.

4.2 Olga embankment

An embankment was loaded to failure at the Olga site,in Mattagami, in Quebec (Dascal et al. 1972). Thefactor of safety computed was 1.6. Trak et al. (1976)re-analyzed the failure using the concept of undrainedpost-peak strength. However, because of the uncer-tainty of the crust strength, an investigation was carriedout in the crust by Lefebvre et al. (1987). In situ shearbox tests and plate loading tests were performed in the1.2 m thick crust but no tests were done in the transi-tion zone which extended to about 3 m depth. A large

number of the vane strength profiles were performedboth in two test trenchs (east and north) and there wassubstantial variability of the vane strength. The resultsof in situ shear box tests, however, were quite consis-tent. The ratio of the undrained strength from in situshear box tests to the field vane strength was aboutone quarter and is shown in Table 4.

4.3 Observations on mass strength of crusts

It is apparent from the results of the in situ testsdescribed in the preceding paragraphs and the study onstiff fissured clays at Nanticoke (Lo et al. 1969) and atMaldon (Bishop and Little 1967; also shown in Table4) that the macroscopic structure of fissuring couldreduce the mass strength to about one quarter to one

9


third that of the intact material; the value would dependon the intensity of fissuring at a particular site. As aguideline, the vane strength in the crust of a soft claydeposit should conceivably be reduced to this range.

It is of interest to note that in the planning and exe-cution of the Gloucester Embankment at the NationalTest Site near Ottawa, the impact of the crust of thesensitive clay was recognized by Dr. M.M. Bozozuk(Bozozuk and Leonards 1972) and it was removedprior to the construction of the test embankment.

5 MICROSCOPIC STRUCTURE

5.1 Conceptual view of microstructure

Investigations into the microscopic structures of soilshave been carried out by numerous authors (see e.g.Mitchell 1976, Rosenquist 1966). As early as 1966,Quigley andThompson (1966) using the X-ray diffrac-tion technique showed that for a block sample of LedaClay, soil fabric underwent a large change once the pre-consolidation pressure as determined in an oedometertest was exceeded. It was hypothesized that cemen-tation bonding was predominantly destroyed at yieldand greater anisotropic loading led to an increasedparallel arrangement of clay particles in the oedome-ter tests. More recently, Leroueil and Vaughan (1990)reviewed the strength behaviour of many natural soilsand weak rocks and considered that the effects ofstructure (microstructure) on engineering behaviourshould be treated as a basic concept in geotechnicalengineering.

A conceptual view of the microstructure of clays isshown in Figure 13. The structure, consisting of thefabric and the cementation bonds, was developed dur-ing and after deposition of the soil under a field stresssystem and physico-chemical environment. The fabricof sensitive clays may be conceived as a highly com-plex space frame and derives its resistance to shearby displacements and deformations of its constituentmembers and joints.The cementation bonds at the con-tacts of clay platelets are randomly distributed, andare brittle in behaviour requiring little deformation torupture. For a given physico-chemical system, the rel-ative contribution of the bonds and fabric to the overallmobilized resistance of the soil to deformation wouldpredominantly depend on the intensity and strength ofthe cementation bonds.

Starting from an equilibrium state, an increase inapplied stresses will be transmitted through the soilskeleton (fabric) producing the deformations arisingfrom (a) the elastic deformation of the soil skeleton,(b) deformation and sliding at points of contact, and(c) deformation of the soil particles. Component (c)may be neglected since the compressibility of the soilskeleton is orders of magnitude greater than that of the

Figure 13. Conceptual view of change of microstructurewith shearing in sensitive clays.

soil particles. The vectoral summation of these micro-scopic deformations are observed as strain in a givendirection.

As the applied stresses are increased, the exter-nal stresses are transferred to the points of contact.Since there is a lack of symmetry in the fabric andthe distribution of bonds, the distribution of normaland shear forces at the contact points is not uniform.In addition, distortion of the soil fabric would inducetensile stress in some contact points. The criteria ofrupture, whether in shear or in tension, will be satis-fied at some contact points leading to bond breakage.The failure at points of contact leads to some particlere-arrangement (see Figure 13), observed externallyas plastic (irrecoverable) deformation. The stressesoriginally carried at the contact points will partly betransferred to the pore water, increasing the pore pres-sure and partly to the neighbouring points of contact.The shearing resistance of the broken contacts wouldreduce to that similar to the post-peak strength of theclay. Therefore, even at external stresses well belowmacroscopic failure of a test specimen, bond break-age occurs and produces some plastic deformationand slight re-arrangement of soil fabric as shown inFigure 13.

This process was well illustrated by incrementalstress-controlled CIU tests on normally-consolidatedsensitive clays in which both plastic deformation(creep) and pore pressure at a constant applied stressincreased simultaneously with time (Lo 1961). Theprogressive nature of bond rupture during shear canalso be illustrated by Figure 14 in which the modu-lus of deformation of St. Louis Clay in CIU and CIDtests are plotted against consolidation pressure. It canbe seen that at half of the failure stress, the trend of

10



Figure 14. Variation of modulus of deformation with con-solidation pressure for St. Louis Clay (after Lo and Morin1972).

modulus variation with consolidation pressure reflectsthat of the curved strength envelope shown in Figure 3.

As the applied stress increases in a triaxial testtowards the peak stress, localization of deformationoccurs due to the formation of a failure zone. Withinthe failure zone at the peak stress, the bond strength isfully mobilized. The test specimen then softens andexhibits a decrease in strength with further strain (morecorrectly, further displacement in the failure zone).Thepost-peak strength is reached at a moderate nominalstrain in the range of 6%10% in sensitive clays. How-ever, particle parallelism can only be approached atmuch larger displacement in the region of the residualstrength in natural clays.

5.2 Laboratory study of cementation bonds

Studies on the source and nature of cementation bondsin sensitive clays in Eastern Canada have been under-taken by Kenny et al. (1967), Yong et al. (1979)and Quigley (1980), among others. While there wassome difference of opinion regarding the details ofthe methods of these mineralogical and geochemicalinvestigations, there appears to be a general agree-ment that calcium carbonate and amorphous materialsincluding SiO2, Fe2O3 and Al2O3 are the most likelycementing agents in these sensitive clays.

To proceed from qualitative to more quantitativeassessment of the contribution of cementation bonds tothe overall shear strength of soft clays, one difficulty isthe lack of baseline reference for natural clays. It seemsappropriate therefore to artificially induce cementationbonds by employing only one potential cementationagent in natural clays, using an untreated sample asa control test throughout the long duration of exper-imentation, so that their contribution to strength canbe ascertained and the possible mechanism of bondingidentified.

5.3 Artificial bonding by electrokinetic process

Following the discussion in the preceding section,the contribution of cementation bonds to the strengthbehaviour of soft clays will firstly be examinedusing artificial bonding achieved by electrokinetic pro-cesses. The bonding agent will be iron compoundsderived from the iron electrodes during the treatment.

The soft clay used in the experiment is a marineclay from Yulchon, South Korea. The liquid limit ofthe clay is 59%, the plasticity index is 27%, and thewater content ranges from 80% to 110%. The clay isnormally consolidated. The undrained shear strengthis between 1 and 6 kPa.

Briefly, the test procedure involved the followingsteps:

(i) Establish the classical relationship of the undrainedshear strength and water content for normally-consolidated clays.

(ii) Set up two identical clay samples under the samepressure and boundary conditions. One sample actsas the control test.

(iii) Treat electrokinetically (EK) the test sample at theapplied voltage of 6.2V using the direct current forseven days, after consolidation at 15 kPa.

(iv) Allow the test to continue for diffusion to take placefor a further 45 days after EK treatment.

The test set-up for EK treatment of theYulchon Clayis shown in Figure 15. Details of the test procedurehave been presented in Micic et al. (2002).

Tests were performed before and after the elec-trokinetic treatment to investigate the changes in thephysical, mechanical and chemical properties of theYulchon Clay due to electrokinetic treatment. The test-ing program included undrained shear strength andwater content measurements, soil chemistry analyses(x-ray fluorescence or XRF, specific surface and cationexchange capacity) and soil surface analyses using ascanning electron microscope (SEM) including energydispersive x-ray (EDX) analyses for identification ofthe elemental composition of the soil. Based on theresults of the tests, the contribution of cementationbonds to the strength behaviour of the Yulchon Claywas evaluated.

5.4 Results of artificially-induced bonding

Analyses of the relationship between the undrainedshear strength and water content of the normally-consolidated (7-15 days of consolidation) YulchonClay show that the undrained shear strength and watercontent of Yulchon soil yield an exponential relation-ship as shown in Figure 16. Results of isotropically-consolidated undrained triaxial (CIU) tests shown inFigure 17 indicate the ratio su/pcof 0.3, where p isthe consolidation pressure. This value is similar to thein situ value of su/p = 0.26 at theYulchon site in South

11



Figure 15. Experimental apparatus (All dimensions in mm;not to scale).

su = 551e-0.05w

0.1

1

10

100

50 60 70 80 90 100 110 120 130

Water Content (w), %

Und

rain

ed S

hear

Str

engt

h (s

u), k

Pa

p' - Consolidation pressureUU - Unconfined compression testCIU - Isotropically consolidated undrained triaxial test

Vane tests

UUp'=15 kPa)CIU(p'=15 kPa)CIU(p'=30 kPa)CIU(p'=45 kPa)

Figure 16. Undrained shear strength of untreated Yulchonsoil versus water content

Korea reported by Hyundai Engineering and Con-struction Co. Ltd. (HDEC) in 1996. As expected, thestress-strain curves in Figure 17 showed no post-peakdecrease in strength.

0

5

10

15

20

25

30

35

40

0 2 4 6 8 10 12 14 16 18 20 22

Axial Strain , %

1-

3, k

Pa

p'=30 kPa

p'=45 kPa

yield

2suyield

Shear Strength vs ConsolidationPressure

0

10

20

30

40

0 10 20 30 40 50p', kPa

s u, k

Pa

p'su su/p'=0.3

Figure 17. Results of CIU triaxial tests on natural YulchonClay

The results of undrained shear strength changesafter EK treatment of the Yulchon Clay are shown inFigure 18. Figures 18(a) and (b) present the relation-ship between the water content and undrained shearstrength after electrokinetic treatment and diffusionphases in the vicinities of the anodes and cathodes,respectively. The change in strength may be attributedto the processes operating in the tests, including:

(a) aging a process of bond growth with time with-out introduction of external agents (Leonards andRamiah 1959, Bjerrum and Lo 1963);

(b) electroosmotic consolidation a process of elec-trically induced water flow from anode to cathode(see e.g. Casagrande 1949, Mitchell and Wan 1977,Lo and Ho 1991); and

(c) deposition of cementation bonds under ionic diffu-sion.

The small increase in strength in the control samplesafter 52 days may be attributed to the process of agingunder the constant applied stress of 15 kPa. Duringelectrokinetic treatment, all three processes would beoperating but the dominant mechanism is electroos-mosis as can be seen by the large decrease in watercontent at the anode region and little change in watercontent at the cathode region. Finally, after the currentis switched off, the mechanism operating would be

12



1

10

100

40 50 60 70 80 90 100 110 120


Und

rain

ed S

hear

Str

engt

h (s

u), k

Pa

(pc=15 kPa; Uo=6.2 V)

After diffusion (45 days)

After EK treatment(7 days)

After 52 days

EK treated-immediately after EK traetment of 7 daysEK traeted-45 days after EK treatment (diffusion)Control-after 52 days

(a) Anode Region

1

10

100

40 50 60 70 80 90 100 110 120


Und

rain

ed S

hear

Str

engt

h (s

u), k

Pa

(pc=15 kPa; Uo=6.2 V)

After diffusion (45 days)

After EK treatment(7 days)

After 52 days

(b) Cathode Region

Figure 18. Undrained shear strength versus water content:(a) at anode and (b) at cathode

ionic diffusion with a small contribution from aging.During this period, deposition of cementation bondspredominately occurs.

Figure 19 illustrates the development of strengthduring the entire experiment by following the strength-water content paths starting from an initial watercontent of 95%.At the anode region, the shear strengthincreased from 4.5 kPa to 16.5 kPa immediately afterelectrokinetic treatment along with a decrease inwater content from 95% to 74%. The undrained shearstrength further increased from 16.5 kPa to 21 kPa aftera diffusion phase of 45 days in spite of an increase inthe soil water content from 74% to 85%. At the watercontent of 85%, consolidation alone as indicated bythe results of the control test would yield a strengthvalue of 7 kPa. Thus, the strength contribution frombonding amounts to 67% of the total strength.

At the cathode region, the undrained shear strengthincreased from 4.5 kPa to 11.5 kPa immediately afterelectrokinetic treatment along with a decrease in watercontent from 95% to 91%. The shear strength fur-ther increases from 11.5 kPa to 15 kPa after 45 daysof the diffusion phase along with a slight decrease inwater content from 91% to 87%. At a water contentof 87%, consolidation only would yield an undrained

1

10

100

65 70 75 80 85 90 95 100 105


Und

rain

ed S

hear

Str

engt

h (s

u), k

Pa Anode

Cathode

Anode

EK

DEKD

Cathode

Untreatedsoil

Untreated Yulchon clay(normally consolidated)

su=551e-0.05w

Figure 19. Development of the undrained shear strength andwater content changes of the Yulchon Clay during and afterEK treatment.

0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

0 2 4 6 8 10 12 14 16 18 20 22

Axial Strain, %

Axi

al S

tres

s, k

Pa

anode region- sample 1

anode region - sample 2

cathode region - sample 1

untreated

cathode region - sample 2

Figure 20. Results of unconfined compression tests.

strength of 5.4 kPa. Thus, the strength from bondingwould constitute 64% of the total strength.

The stress-strain curves from unconfined compres-sion tests on treated soil are presented in Figure 20. Ascan be seen, the results of compression tests are con-sistent with the results of vane tests discussed earlier,showing that the undrained shear strength increaseddue to EK treatment. In addition, brittleness devel-oped in the soil as a result of electro-cementation. It isalso noted that the brittleness is more prominent at thecathode than at the anode region, which is consistentwith the strength development paths in Figure 19.

13


1.5

1.6

1.7

1.8

1.9

2

2.1

2.2

2.3

2.4

1 10 100 1000

Applied Pressure, kPa

Voi

d R

atio

pc'=40 kPa

pc'=16 kPa pc'=42 kPa

Cathode

Anode

Figure 21. Results of consolidation tests on Yulchon Clay.

Oedometer tests were performed on specimensfrom the anode and cathode region as well as from thecontrol test. The results are shown in Figure 21. It maybe seen that a preconsolidation pressure of approxi-mately 40 kPa has developed in both the anode andcathode region as a result of electro-cementation. Thecontrol test gives a preconsolidation pressure of 16 kPacompared with the applied pressure of 15 kPa. It maytherefore be observed that an overconsolidation ratioof about 2.5 has been induced by cementation bonding.

The mechanism of this electro-cementation may beattributed to selective sorption and ionic exchange ofionic species on clay particle surfaces and precipita-tion of amorphous chemical compounds such as ironoxide/hydroxide and calcium carbonate which serveas cementation agents (Quigley 1980). X-ray fluores-cence (XRF), specific surface and cation exchangecapacity (CEC) analyses were performed on the soilsamples to detect the chemical changes in the soil dueto electrokinetic treatment and to identify cementingagent(s) involved.The XRF analyses provide the majorelement composition of the soil. The results of theanalyses shown in Table 5 show that the percentageof iron oxide (Fe2O3) increased significantly in thesoil after electrokinetic treatment while the percent-ages of other oxides (e.g. SiO2, TiO2, Al2O3, MnO,MgO, CaO, K2O, Na2O, P2O5, Cr2O3) only slightlychanged. In particular, the percentage of iron oxideincreased from 5.7% up to 11.8% while the percentageof other potential bonding agents of SiO2 and Al2O3showed no increase. The increase in iron oxides is alsoconfirmed by the change in the soil colour from greyto yellowish-brown in the zone of influence of elec-trokinetic treatment. The source of the iron was fromthe steel electrodes, which corroded during the elec-trokinetic treatment. The released iron precipitated asoxide or hydroxide due to the extremely low solubility

Table 5. Results of XRF analyses of Yulchon Clay.

Oxides (%) Control EK Treated Soil

SiO2 56.55 49.90TiO2 0.72 0.64Al2O3 16.70 14.61Fe2O3 5.74 11.78MnO 0.09 0.13MgO 2.41 1.82CaO 1.23 2.50K2O 2.96 2.60Na2O 1.67 1.39P2O5 0.10 0.95Cr2O3 0.01 0.02L.O.I. 11.60 13.10Total 99.78 99.45

Table 6. Results of post-treatment chemical tests.

Properties Control EK Treated Soil

Iron Oxide Fe2O3 (%) 5.7 11.8Specific Surface (m2/g) 23 34CEC (meq/100g soil) 6.7 26.4Iron Fe (Wt%) 8 36

of iron in the normal pH range of soils. The ironoxide that adsorbed on soil particle surfaces induced acementation effect that led to the consequent develop-ment of strong aggregation of soil particles and thusan increase in the soil shear strength.

The results of specific surface and CEC analysesof the treated soil are listed in Table 6. For compari-son, the corresponding values of untreated soil are alsoincluded in the table. It can be seen from the table thatthe values of specific surface and CEC of the electroki-netically treated soil particles were higher than those ofuntreated soil. This increase in specific surface area,and thus in the CEC, also indicates the presence ofthe higher content of iron oxides in the treated soilbecause it is known that iron oxides have high spe-cific surface area amenable to act as coating on otherparticles (Dixon et al. 1977).

In addition, Energy Dispersive X-ray (EDX) anal-yses were performed to identify the elemental com-position of the soil. The average of the percentage ofiron per total weight of the untreated soil was approx-imately 8 Wt%, while the percentage of iron aftertreatment was about 36 Wt%.

The microscopic structure of the Yulchon Claybefore and after EK treatment was studied using ascanning electron microscope (SEM). The SEM anal-yses were undertaken in order to visually identify theoccurrence of cementation in the soil due to elec-trokinetic treatment. Figures 22(a) and (b) show thesurfaces of the untreated (control) and treated soils,

14


Figure 22. Electron microscopy images of Yulchon Clay:(a) Control samples; (b) EK treated samples.

respectively. It is evident that some amorphous cemen-tation compound(s) were formed and precipitated onthe clay particles.

Finally, it is noted that the iron oxide (Fe2O3) hasbeen measured in natural St. Alban and Gatineau Clay(Yong et al. 1979) with values of 5% and 6%, respec-tively. These values are comparable to that of YulchonClay used in the experiments as shown in Table 6. Inaddition, the authors suggested that the oxides wouldcoat the particles. The EM image in Figure 22 lendssupport to this hypothesis.

From this study on electrokinetically inducedcementation bonds, the following observations maybe made:

(1) Iron oxides can act as an effective cementing agentin soft clays.

(2) Cementation bonds can contribute up to approxi-mately 60% to 70% of the undrained strength ofthe clay with brittle behaviour.

(3) Similarly, an overconsolidation ratio of about 2.5can be induced by electro-cementation.

Figure 23. CIU-OC tests, i= 0, Initial c = 210 kPa, St.Vallier Clay.

5.5 Estimate of bond strength in some naturalmaterials

Although the existence of bonds in soft clays hasbeen accepted by some researchers for some time (e.g.Crawford 1963, Kenney et al. 1967), direct measure-ment of bond strength in natural soils is difficult andtheir order of magnitude can only be inferred. In thecase of St. Vallier Clay, the drained tensile strengthis only about 3 kPa. This would represent the mini-mum bond strength under tensile stress induced at thecontact points.

In an attempt to evaluate the bond strength undershear, three series of CIU tests were performed on St.Vallier Clay by isotropically consolidating specimenstrimmed from block samples to pressures of 140, 210and 280 kPa and then reducing the consolidation pres-sure to achieve OCRs up to eight (see Morin 1975).The results of one of the series are shown in Figure 23in which the post-peak envelope from Figure 8 is alsoshown. It may be observed that for OCR exceedingthree, the shear strengths lie close to the post-peakenvelope but not on the extension of the unstructuredenvelope. Similar observations may be made on resultsfrom the St. Louis Clay. The results of these tests arean additional indication of the robustness of the post-peak envelope. Using this envelope as the baselinereference, the maximum bond strength under shear forSt. Vallier Clay would be about 20 kPa and representsabout 30% of the shear strength in the effective stressregion considered. Similar results were also obtainedfor St. Louis Clay.

Substantially higher bond strength may exist in stiffquick clays in the lower St. Lawrence region. The soilinvolved in the Toulnustouc Slide (Conlon 1966) hasa liquid limit of 22, plasticity index of 4, with a highliquidity index of 3.4. The undrained shear strengthis 400 kPa. A drained tension test indicated that theminimum tensile bond strength is about 17 kPa. Thebond strength in shear may be interpreted to be asmuch as 350 kPa.

15



Quigley (1968) performed a mineralogical analy-sis on a small block sample of the clay and reportedthe strong bonding exhibited by the Toulnustouc Clayis also related to aluminium and iron hydroxide pre-cipitates in the soil. These materials probably formbonds in two ways: (1) by direct precipitation to forma cement linking the soil grains together, and (2) bygrowing in the mineralogical continuity at the edgesof the clay crystals, thus increasing their size. Thelatter would result in increased Van der Waals attrac-tive forces as crystals grow closer together and couldeven form cementation bonds if the crystals came intocontact with one another. The reasonableness of thishypothesis has been supported by the results of theartificial cementation study in Section 5.4.

An example of very large bond strength is describedin Leroueil and Vaughan (1990) for a mudstone inJapan (Ohtsuki et al. 1981). An examination of theirdata shows that in the normal stress range from 1500to 3000 kPa, the friction angle is only 8 with theshear strength of about 1800 kPa. In this stress range,the shear strength mobilized is therefore mostly bondstrength.

It can be observed from these cases that while thebond strength in tension is low, the bond strength inshear of natural material may differ by three orders ofmagnitude and may constitute the major component ofthe total shearing resistance that are measured in con-ventional tests in some natural materials. The degreeto which it can be mobilized depends on the nature ofthe engineering problem under consideration.

6 ANALYSIS OF THE VERNONEMBANKMENT

6.1 Brief description of failures

A dramatic and most instructive case history waspresented by Crawford et al. (1995) who describedtwo consecutive failures of an embankment on softclay, in spite of the fact that two test embankmentswere already constructed on either side of the failuresand that the test embankments were higher than theembankments that failed. The site is at Vernon, BritishColumbia. The subsoil conditions shown in Figure 24consisted of approximately 4 m of interlayered sand,silt and clay, followed by a 5 m thick stiff to very stiffclay crust, then by a deep deposit of soft to firm siltyclay. The undrained shear strengths measured by fieldvane tests were approximately 80 kPa in the stiff clayand 30 to 40 kPa in the soft to firm clay. The plasticityindex was about 35 and the natural moisture contentsvaried from 60% to 80%. Figure 25 shows the loca-tions of the test embankments and the embankmentthat failed twice. The west test embankment was con-structed to approximately 11.5 m thickness. The easttest embankment with wick drains was constructed

to 12 m. The test fills were well instrumented. Porepressures and settlements were measured during theconstruction of the test fills and the road embankment.

Construction of the road embankment started inearly December 1988 and slowly filled to 7 m to 9.5 malong the alignment by June 30, 1989, when the firstfailure occurred on the north side encompassing a por-tion of the east test fill, as shown in Figures 25 and 26.The test fill had been in place since 1986, and accordingto the results of monitoring, all excess pore pressureshad dissipated (see Fig. 12 of Crawford et al. 1992).The failure was deep-seated and probably circular. Itappears that the only significant warning sign was thatthe ratio of the pore pressure increase to the appliedloading increase u/p approached one within thefailure zone. The pore pressure response to embank-ment load within the failure area is shown in Figure27.

Reconstruction of the embankment was carried outby adding 5 m thick and approximately 30 m wideberms on both sides of the failure. Filling startedin August 1989 and progressed at a very slow rate.On March 10, 1990, a second failure, much larger inextent and including most of the first failure, occurredbetween the two test fills that had been in existencesince 1986 (Figures 25 and 26). The height of the fillat the time of the second failure was 11.2m, which issomewhat higher than that of the first failure. The roadwas eventually completed with berms and lightweightfill.

This case record, with test fills and well executedinstrumentation and monitoring, led to several obviousbut perplexing issues.

1. Why was the observational approach, which is gen-erally accepted and now a time-honoured method,not successful in preventing either the first orsecond failure?

2. In what way are the results of the two test fillsmisleading? Is the degree of natural horizontal vari-ation of soil properties sufficient to cause the resultsof the test fills to be inapplicable?

3. Why did the designed provision of berms notprevent the second failure?

To investigate these issues, a series of limit equi-librium and finite element analyses were performedand the results of these analyses are discussed in thefollowing sections.

6.2 Limit equilibrium analyses

Crawford et al. (1995) performed stability analyses forthe first failure assuming a uniform undrained strengthin both the crust and in the soft clay layer below thecrust. The results of their study showed that a factorof safety of approximately one could be obtained fora crust strength of 50 kPa and a strength of 30 kPa inthe soft clay layer below the crust.

16


Figure 24. Profiles of water contents, Atterberg limits and shear strengths (after Crawford et al. 1995)

Figure 25. Site plan showing location of test fills and failure zones (after Crawford et al. 1995)

A more detailed representation of the subsoilstrength profiles was used in this paper based on thevane strength data shown in Figure 24. In accordancewith the observations on the effect of fissures on themass strength discussed in Section 4, the strength ofthe crust was corrected to 40 kPa down to a depth of6 m where the strength decreases in the transition zone

to 9 m depth from which the strength increases lin-early with depth. Bearing in mind that the depth ofthe slip surface lay within the first 15 m depth, threestrength profiles are shown in Figure 28, together withthe 1960 and 1985 measured vane strength. It is con-sidered that the middle profile marked M appears tobe the most representative of the vane strength data

17



Figure 26. Longitudinal section through the embankment (after Crawford et al. 1995)

Figure 27. Height of fill, settlement, and piezometric surface at centre line of station 27+80 duringconstruction (after Crawford et al. 1995)

18



while the higher (by 20%) strength profile markedH and the slightly lower (by 10%) strength profilemarked L are obviously within reasonable limits ofinterpretation of the strength data.The material param-eters of the fill used were unit weight of 20.4 kN/m3

and c = 0, = 33, as in Crawford et al. (1995).Table 7 summarizes the material parameters used inthe analysis.

The results of stability analysis are shown inTable 8.It may be seen that, for the first failure, both the M-profile and the L-profile yield a factor of safety not farfrom one. As discussed earlier, both profiles are withinreasonable limits of interpretation of measured vanestrength data. Without correcting the crust strength toaccount for fissures, the factor of safety would be 1.3.The factors of safety for the second failure are slightlyhigher than the corresponding ones of the first failurebut are still within the limits of reliability of = 0analysis. It is recognized that part of the fill would havesettled into the subsoil rendering the results somewhatdifficult to interpret. Nonetheless, the results of thesecond failure may be considered as supplementary

Undrained Strength, Su (kPa)0 20 40 60 80 100 120 140

Dep

th (

m)

-60

-40

-20

0

20

H Strength ProfileM Strength ProfileL Strength ProfileMeasured Vane Strength 1985Measured Vane Strength 1960

-6-9

Fill

Crust

Transition Layer

Clay Layer

Figure 28. Distribution of vane strength with depth (adaptedfrom Crawford et al. 1995)

Table 7. Material properties used in the limit equilibrium analysis of the VernonEmbankment.

Depth Strength Profile (kPa) Unit Weight,(m) (kN/m3)

H M L

Crust Layer 06 40 40 40 20Transition Layer 69 4035 4028 4024.5 17Soft-Stiff Clay Layer 940 3575 2860 24.552.5 17

evidence, which is consistent with results of analysisof the first failure.

From the discussions in the preceding paragraphs,it is apparent that the instability condition of the Ver-non Embankment is similar to other embankments insoft to firm sensitive clays. While limit equilibriumanalysis might have (from hindsight) predicted theinstability of the two failures, conventional stabilityanalysis alone would not have addressed the questionsin Section 6.1.

6.3 Finite element analysis of vernon embankment

It has been recognized that the development of an over-stressed zone (plastic region) in soft clay controls thedevelopment of high pore pressures and thus the sta-bility of embankments with low factors of safety (Lo1973; Law 1975). In order to explore the behaviour ofthe Vernon Embankment in more detail, finite elementanalyses were performed.

6.3.1 Method of analysisThe first series of analyses carried out involved elasto-plastic total stress analysis under plane strain conditionusing the program AFENA (Carter and Balaam 1995)for the two successive failures. The parameters usedare the same as those used in the limit equilibrium anal-ysis (Table 7). Additional parameters required are thecoefficient of earth pressure at rest, Ko, which is takento be 1.04 in the crust and 0.84 in the soft clay. Theundrained elastic modulus, Eu, was evaluated assum-ing an Eu/Su ratio of 500 and Su from the M-profile inFigure 28. The fill strength used was c = 10 kPa and = 33.

The construction of the embankment was simulatedby activating the elements of fill material layer by layer,

Table 8. Factor of safety with different strength profiles.

The First failure The Second failure

FS FS

H Strength Profile 1.19 1.29M Strength Profile 1.07 1.13L Strength Profile 1.00 1.04

19


Unit: Meter

26.7

55.0

The centre line of the embankment

11.5

9.411.2

9.3

1.0

1.5

5.0

7.5

1.5

1.0

Step10 (0.9m)Step9 (1.1m)Step8 (1.1m)Step7 (1.1m)Step6 (1.1m)Step5 (1.0m)Step4 (1.0m)Step3 (1.0m)Step2 (1.0m)Step1 (1.0m)

Step11 (0.9m)

Figure 29. The numerical construction scheme (The shaded area represents the construction after the first failure).

80Soom

th B

ound

ary

Smoo

th B

ound

ary

300Rigid Boundary

Unit:meter

Figure 30. FEM mesh and boundary conditions

nine layers for the first failure and eleven layers for thesecond layer as shown in Figure 29. The mesh used isshown in Figure 30.

6.3.2 Results of analysisThe incremental simulation of the embankment con-struction portrays the development of the plastic zoneand velocity field. Figure 31 illustrates the extent ofthe plastic zone and velocity field at an embankmentheight of 8.3 m (prior to failure) and 9.4 m (at fail-ure), respectively. The distinct changes in the plasticzone and velocity field when the fill height reached9.4 m can be observed. As the embankment heightapproaches the collapse load, the plastic region extendsto the ground surface outside the embankment and akinematic collapse mechanism develops as shown inthe velocity field. At 9.4 m, both the plastic zone andvelocity field indicate a failure state is imminent or hasbeen reached.

The propagation of the plastic zone with an increasein embankment height is shown in Figure 32 togetherwith the velocity field boundary. The measured lateraldeflection close to the toe at Station 27+ 80 (Figure 11,

Crawford et al. 1995) is also shown. From this figure,the following observations may be made:

(a) The plastic zone starts to form in the soft claybelow the crust and engulfs the location of thepiezometer at 10 m depth when the embankmentheight reaches 4 m at Stn. 27+ 80. Subsequent tothe yielding of the soil at this moment, an increasein rate of pore pressure rise may be expected. Fig-ure 33 shows the measured pore pressure with anincrease in embankment height. It can be seen thatthe yielding of the clay is well indicated by theresults of pore pressure measurements.

(b) The depth and the overall location of the velocityfield boundary are in general agreement with theslip surface deduced by Crawford et al. (1995); and

(c) The horizon of maximum deflection at failureagrees well with the location indicated by theresults of inclinometer measurements.

The effect of strength profiles on the prediction ofthe critical height of the embankment is shown in Fig-ure 34. It can be seen that the H-profile over predicts,the L-profile under predicts slightly and the M-profileyields good agreement with the observed criticalheight of 9.4 m for the first failure. The computedsettlement with embankment height relationships arecompared with the measured settlements in Figure 35.Bearing in mind there would be some effect of partialconsolidation, it can be seen that there is overall con-sistency between the results of the M-profile and theobserved settlements.

From the discussion in the preceding sections, itis apparent that there is overall general agreementbetween the results of analysis and the observed fieldbehaviour including critical height, pore pressure, lat-eral deflection, settlement and position of the slipsurface.

6.3.3 Results of analysis of second failureSimilar analyses were carried out for the second failureusing the same parameters as for the first failure. The

20


Figure 31. Plastic Zones and Velocity Fields at embankment heights of H= 8.3 m and H= 9.4 m

Distance from the centre line, m

0 20 40 60 80 100

0 20 40 60 80 100

Dep

th, m

-80

-70

-60

-50

-40

-30

-20

-10

0

10

20

-6m

-9m

Calculated Velocity Field Boundary (H=9.4m)

Measured Lateral Defection

Loca

tion

of th

e pi

ezom

eter

H=3.7m

4.0

5.06.1

8.37.2

9.4

Crawford's Deduced Slip Surface

Su=40kPa

Su=28kPa

M-Profile

Figure 32. Development of the plastic zone in the foundation at increasing embankment height

21



Embankment Height,m

0 1 3 4 5 6 7 8 9 10

PW

P, m

0

2

4

6

8

10

12

14

16

18

20

The beginning of yieldingin FEM analysis

2

Figure 33. Measured pore water pressure on 10 m depth at centre line of station 27+80.

Embankment Heigh (m)

0 1 2 3 4 5 6 7 8 9 10 11 12

Set

tlem

ent a

t Cen

tre

Line

of E

mba

nkm

ent (

m)

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

Calculated settlement with H strength profileCalculated settlement with M strength profileCalculated settlement with L strength profile

Observed critical heightof the first failure=9.4m

Figure 34. Settlement of embankment centre vs embank-ment height with different strength profiles at station 27+ 80(the first failure).

results of computed embankment height versus settle-ment relationships for the three strength profiles areshown in Figure 36. The results for the L-profile yieldagreement with the observed critical height of 11.2 m.One interpretation would be that this might indicatean overall loss of strength of about 10% after the firstfailure due to disturbance. This interpretation would

Embankment Heigh,m

0 4 10 11 12

Set

tlem

ent a

t Cen

tre

Line

of E

mba

nkm

ent,

m

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

Calculated settlement with H strength profile Calculated settlement with M strength profileCalculated settlement with L strength profileMeasured settlement

9.4321 98765

Figure 35. Measured and calculated settlements of theembankment centre station 27+ 80 (the first failure).

be consistent with the factors of safety computed inTable 8.

6.3.4 Analysis of waterline test fillTwo test fills were successfully constructed on thewest (Waterline Test Fill) and east (West AbutmentTest Fill) of the two failures as shown in Figure 25.Because the performance of the West Abutment TestFill was affected by the installation of prefabricated

22


Embankment Heigh,m

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Set

tlem

ent a

t Cen

tre

Line

of E

mba

nkm

ent,

m

-1.6

-1.4

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

Calculated settlement with H strength profileCalculated settlement with M strength profileCalculated settlement with L strength profile

Observed critical height ofthe second failure=11.2m

Figure 36. Vertical displacement of embankment centre vsembankment height with different strength profiles at station27+ 80 (the second failure).

vertical (wick) drains, only the Waterline Test Fillwill be analyzed so as to investigate the differencein behaviour between the failed embankment and thestable condition of the Waterline Test Fill.

An examination of the geometries of the WaterlineTest Fill shows that the problem is closer to three-dimensional than a plane strain condition. Therefore,any plane strain analysis (including limit equilibriumanalysis) based on plane strain conditions may be mis-leading. Although a 3-D elastoplastic analysis wouldbe preferable, a simpler axi-symmetric analysis wasperformed so as to obtain some insight, as a firstapproximation, into the impact of geometry on thevast difference in behaviour of the embankments. Therectangular geometry of the Waterline Test Fill wasidealized to a circular load with its diameter equal tothe average dimension of the two sides.

Figure 37 shows the progress of the plastic zonefrom 9 m to 11.2 m to which the test fill was suc-cessfully completed. It is evident that at 11.2 m, thecondition is that of a contained plastic zone and thetest fill is stable. (A conventional limit equilibriumanalysis with the M-profile would have shown that thefactor of safety would be well below unity. In con-trast, a back analysis assuming a factor of safety ofone would have indicated high strength. Both resultswould be misleading.)

The plastic zones at H= 9.0 m and H= 9.4 m for astrip and circular embankment are shown in Figure 38.The large difference in extent of the plastic zones dueto different geometries of loading is evident. In addi-tion, the propagation from 9.0 to 9.4 m is quite smallfor the circular load. In contrast, the increment 0.4 mof loading for the strip embankment results in a con-tinuous plastic zone that has propagated to the groundsurface leading to collapse.

It is therefore suggested that observations at theWaterline Test Fill may not be directly applicable to

Distance from the centre line, m0 20 40 60 80 100

0 20 40 60 80 100

Dep

th, m

-80

-70

-60

-50

-40

-30

-20

-10

0

10

20

H=9.00m Waterline test fill (Axisymmetric Strain)H=9.40m Waterline test fill (Axisymmetric Strain)H=11.4m Waterline test fill (Axisymmetric Strain)

Equivalent Axisymmetric Geometry of Waterline Test FillH=11.4m

1.01.4

-6m-9m

Su=40kPa

Su=28kPa

M-Profile

Figure 37. Development of plastic zone underWaterline testfill (Axi-symmetric strain assumption).

0 20 40 60 80

Distance from the centre line, m100

Dep

th, m

-80

-70

-60

-50

-40

-30

-20

-10

0

10

20Plane Strain H=9.0m (Strip Embankment)Plane Strain H=9.4m (Strip Embankment)Axisymetric Strain H=9.0m (Waterline Test Fill)Axisymetric Strain H=9.4m (Waterline Test Fill)

-6m-9mFailure

1.51.0

1.41.0

Equivalent Axisymmetric Geometry of Waterline Test Fill

Cross-section of Road Embankment on Station 27+80

Su=40kPa

Su=28kPa

M-Profile

Figure 38. Development of plastic zones under Waterlinetest fill and strip embankment at station 27+ 80.

the road embankment due to the difference in devel-opment of the plastic zone under different loadingconfigurations.

In a subsequent section, case records of well definedloading geometries will be analyzed to verify thefindings discussed for the Vernon case records.

23


7 EFFECT OF EMBANKMENT GEOMETRY ONBEHAVIOUR OF FOUNDATION CLAY

This section examines the Sk-Edeby Test Field inorder to verify the effect of embankment geometryon the behaviour of underlying soft clay deposits asseen in the Vernon embankment and test fills.

The Sk-Edeby case involved construction andmonitoring of embankments with well defi

dave chan_ international conference on soft soil engineering. _4, 2006, vancouver, british...

Documents